Prerocessing Repo Additions (#68)

* updated

* lock

* updated mineral sites

* here

* latest

* updated

* multiple vector rasters

* latest

* latest

* format

* list of processed ids

* version

cdr_schemas/cdr_responses/prospectivity.py

@@ -1,11 +1,13 @@
 from datetime import datetime
 from typing import List, Union
 
+from geojson_pydantic import LineString, Point, Polygon
 from pydantic import BaseModel, ConfigDict, Field
 
 from cdr_schemas.prospectivity_input import (
     CreateCriticalMineralAssessment,
     CreateDataSource,
+    DataTypeId,
     TranformMethods,
 )
 from cdr_schemas.prospectivity_models import NeuralNetUserOptions, SOMTrainConfig
@@ -26,8 +28,35 @@ class CreateProcessDataLayer(BaseModel):
     data_source: DataSource = Field(description="Data source to create this layer")
     title: str = Field(description="Title to use for processed layer")
     transform_methods: TranformMethods = Field(
-        default="", description="Transformation method used"
+        default_factory=list, description="Transformation method used"
     )
+    label_raster: bool = Field(description="A label layer for training")
+
+
+class DataTypeIdWithGeom(DataTypeId):
+    geom: Point | LineString | Polygon = Field(description="Adding feature coords")
+
+
+class CreateVectorProcessDataLayer(BaseModel):
+    label_raster: bool = Field(
+        default=False, description="Layer used to train prospectivity models"
+    )
+    title: str = Field(description="Title to use for processed layer")
+    evidence_features: List[DataTypeIdWithGeom] = Field(
+        default_factory=list, description="Feature ids from the cdr"
+    )
+    extra_geometries: List[Point | LineString | Polygon] = Field(
+        default_factory=list,
+        description="site locations selected by expert. Use EPSG:4326 only",
+    )
+    transform_methods: TranformMethods = Field(
+        default_factory=list, description="Transformation method used"
+    )
+
+
+class ProcessedDataLayer(BaseModel):
+    layer_id: str = Field(description="Layer id")
+    download_url: str = Field(description="Download url")
 
 
 class ProspectModelMetaData(BaseModel):
@@ -40,6 +69,22 @@ class ProspectModelMetaData(BaseModel):
     cma: CriticalMineralAssessment = Field(description="CMA info")
     model_type: str
     train_config: Union[SOMTrainConfig, NeuralNetUserOptions]
-    evidence_layers: List[CreateProcessDataLayer]
+    evidence_layers: List[ProcessedDataLayer] = Field(
+        description="Processed data layer ids."
+    )
+
+    model_config = ConfigDict(protected_namespaces=())
+
+
+class ProcessDataLayers(BaseModel):
+    cma: CriticalMineralAssessment = Field(description="CMA info")
+
+    evidence_layers: List[CreateProcessDataLayer] = Field(
+        default_factory=list, description="Datasource and preprocess steps"
+    )
+    vector_layers: List[CreateVectorProcessDataLayer] = Field(
+        default_factory=list,
+        description="Vector features and preprocess steps. EPSG:4326",
+    )
 
     model_config = ConfigDict(protected_namespaces=())

cdr_schemas/prospectivity_input.py

@@ -2,7 +2,7 @@
 from enum import Enum
 from typing import List, Optional, Union
 
-from geojson_pydantic import MultiPolygon
+from geojson_pydantic import LineString, MultiPolygon, Point, Polygon
 from pydantic import BaseModel, ConfigDict, Field
 
 from cdr_schemas.prospectivity_models import (
@@ -102,32 +102,73 @@ class CreateCriticalMineralAssessment(BaseModel):
 # MTRI UI TO CDR:
 # define preprocessing actions
 class DefineProcessDataLayer(BaseModel):
-    cma_id: str = Field(description="ID of the cma")
-    data_source_id: str = Field(description="Data source id used to create this layer")
+    data_source_id: str = Field(
+        description="Processed data source id used to create this layer"
+    )
     title: str = Field(description="Title to use for processed layer")
     transform_methods: TranformMethods = Field(
         default_factory=list, description="Transformation method used"
     )
+    label_raster: bool = Field(
+        default=False, description="Layer used to train prospectivity models"
+    )
 
 
 # TA3 TO CDR:
 # Send along with a processed data layer used for training to support their model output.
 # TA3 can send each layer of the training stack used to generate the output one layer at a time
+
+
+class RawDataType(str, Enum):
+    MINERAL_SITE = "mineral_site"
+    POINT = "point"
+    LINE = "line"
+    POLYGON = "polygon"
+    TIF = "tif"
+    VECTOR = "vector"
+
+
+class DataTypeId(BaseModel):
+    raw_data_type: RawDataType = Field(description="Type of feature.")
+    id: str = Field(description="Id of feature in cdr")
+
+
 class SaveProcessedDataLayer(BaseModel):
-    model_run_id: str = Field(
-        description="Connect this processed data layer to a model run output layer"
-    )
-    data_source_id: str = Field(description="Data source id used to create this layer")
     cma_id: str = Field(description="ID of the cma")
-    title: str = Field(description="Title for processed layer")
+    title: str = Field(description="Title of processed layer")
+    label_raster: bool = Field(
+        default=False, description="Layer used to train prospectivity models"
+    )
+    raw_data_info: List[DataTypeId] = Field(
+        default_factory=list, description="cdr ids and types of all features used"
+    )
+    extra_geometries: List = Field(
+        default_factory=list, description="Extra geometries used to create this layer"
+    )
     system: str
     system_version: str
     transform_methods: TranformMethods = Field(
-        default="", description="Transformation method used"
+        default_factory=list, description="Transformation methods used"
     )
     model_config = ConfigDict(protected_namespaces=())
 
 
+class DefineVectorProcessDataLayer(BaseModel):
+    label_raster: bool = Field(
+        default=False, description="Layer used to train prospectivity models"
+    )
+    title: str = Field(description="Title to use for processed layer")
+    evidence_features: List[DataTypeId] = Field(
+        default_factory=list, description="cdr ids and types of all features used"
+    )
+    extra_geometries: List[Point | LineString | Polygon] = Field(
+        default_factory=list, description="Extra geometries to be used"
+    )
+    transform_methods: TranformMethods = Field(
+        default_factory=list, description="Transformation methods used"
+    )
+
+
 # MTRI UI to CDR:
 # defines the cma, model training config and layer preprocessing steps
 class CreateProspectModelMetaData(BaseModel):
@@ -139,9 +180,28 @@ class CreateProspectModelMetaData(BaseModel):
     organization: str = ""
     model_type: str
     train_config: Union[SOMTrainConfig, NeuralNetUserOptions]
+    evidence_layers: List[str] = Field(
+        description="List of ids of processed data layers"
+    )
+
+    model_config = ConfigDict(protected_namespaces=())
+
+
+# MTRI UI to CDR:
+# defines the layer preprocessing steps
+class CreateProcressDataLayers(BaseModel):
+    cma_id: str = Field(description="CMA id")
+    system: str
+    system_version: str
+
     evidence_layers: List[DefineProcessDataLayer] = Field(
-        description="Datasource and preprocess steps"
+        default_factory=list, description="Datasource and preprocess steps"
     )
+    vector_layers: List[DefineVectorProcessDataLayer] = Field(
+        default_factory=list,
+        description="A list of raster to be created using a set of vector features",
+    )
+
     model_config = ConfigDict(protected_namespaces=())